摘要
利用不可约矩阵和相对不变量理论提出了几种点模式匹配新算法，它们可分别用来解决相似变换和仿射变换下具有相同点数的两个点模式的匹配问题．这些算法的基本出发点是将待匹配的两个二维点集分别转化成为一个ｎ 维空间中的向量（也就是ｎ 维空间中的点），只要对这两个向量中的各元素进行简单的排序就可解决对应的原来点模式的匹配问题．
Point pattern matching is an important problem in the fields of computer vision and pattern recognition. Its task is to pair up the points in two images when there is a geometric transformation relating the two images. In this paper, several new algorithms are proposed to solve the problem of matching two point sets with the same cardinality under an affine transformation or a similarity transformation. These new algorithms are based on irreducible matrix and relative invariant. Their essential idea is transforming the two dimensional point sets with n points into the vectors of n dimension space. When these vectors are considered as one dimensional point patterns, the original point pattern matching problem is changed from two dimensional form to one dimensional form. All these algorithms aim at reducing the point matching problem to that of sorting vectors in n dimension space as long as the sensor noise does not alter the order of the elements in the vectors. Theoretical analysis and simulation results show that the new algorithms are effective.
出处
《计算机学报》
EI
CSCD
北大核心
1999年第7期740-745,共6页
Chinese Journal of Computers
基金
国家教委博士点基金
关键词
点模式匹配
仿射变换
不可约矩阵
图像识别
Point pattern matching, similarity transformation, affine transformation, irreducible matrix.
